J. Biochem. 130. 191-198 (2001)

X-Ray Absorption Spectroscopic Analysis of the High-Spin Ferriheme Site in Substrate-Bound Neuronal Nitric-Oxide Synthase1

Nathaniel J. Cosper,* Robert A. Scott,*,2 Hiroyuki Hori,•õ Takeshi Nishino,•õ and Toshio Iwasaki•õ,2

Department of Chemistry; University of Georgia, Athens. GA 30602-2556, USA; and *Depatment of Biochemistry and Molecular Biology, Nippon Medical School. 1-1-5 Sendogi. Buukyo-ku, Tokyo 113-8602

Received March 27, 2001; accepted April 17, 2001

Nitric oxide synthase (NOS) catalyzes the conversion of L-arginine to citrulline and nitric oxide through two stepwise oxygenation reactions involving NƒÖ-hydroxy-L-argin ine, an -bound intermediate. The NƒÖ-hydroxy-L-arginine- and arginine-bound NOS ferriheme centers show distinct, high-spin electron paramagnetic resonance sig nals. Iron X-ray absorption spectroscopy (XAS) has been used to examine the structure of the ferriheme site in the NƒÖ-hydroxy-L-arginine-bound full-length neuronal NOS in the presence of (6R)-5,6,7,8-tetrahydro-L-biopterin. Iron XAS shows that the high-spin ferriheme sites in the NƒÖ-hydroxy-L-arginine- and arginine-bound forms are strikingly similar, both being coordinated by the and an axial thiolate ligand, with an Fe-S distance of ca. 2.29 A. Cu2+ inhibition slightly affects the spin-state equilibrium, but causes no XAS-detectable changes in the immediate ferriheme coordination environ ment of neuronal NOS. The structure and ligand geometry of the high-spin ferriheme in arginine-bound neuronal NOS are essentially identical to those of the NƒÖ-hydroxy-L-argi nine-bound form and only slightly affected by the divalent cation inhibitor of consitu tive NOS.

Key words: electron paramagnetic resonance spectroscopy, heme, neuronal nitric oxide synthase, nitric oxide synthase, X-ray absorption spectroscopy.

Nitric oxide (NO) serves as a physiological vasodilator, neu ine, and an interdomain binding site for a tetrathiolate zinc rotransmitter, and cytostatic agent in mammalian cells, center (10-13). During NO synthesis, the two flavins (FAD and is generated by a family of termed nitric oxide and FMN) in the reductase domain mediate electron trans synthases (NOSs) (1-5). All known NOS isoforms are each fer from NADPH to the heme group in the oxygenase do catalytically active as a homodimeric, bidomain flavo-heme- main, where oxygen binding and activation take place (1, 6, 14-16). Spectroscopic studies on the heme centers of the protein, consisting of a flavin-containing C-terminal reduc tase domain and a home-containing N-terminal oxygenase NOS isoforms involving absorption, electron paramagnetic domain (6-9). The oxygenase domain also contains binding resonance (EPR), resonance Raman, and magnetic circular sites for 6(R)-5,6,7,8-tetrahydro-L-biopterin (H4BP), L-argin- dichroism techniques, along with mutational analysis, have suggested that an endogenous thiolate sulfur donor ligand is coordinated to the central heme iron, as in the cases of 1 This investigation was supported in part by Grants-in-Aid for Sci entific Research on Priority Areas from the Ministry of Education, P450 and chloroperoxidase (17-26). Science, Sports and Culture of Japan ("Biometallics" 08249104 to Neuronal NOS (nNOS) generates NO through the two- T.N. and "Biomachinery" 11169237 to T.I.), Grants-in-Aid from the step oxygenation reactions that convert L-arginine to citrul Ministry of Education, Science. Sports and Culture of Japan line. Its activity is dependent on oxygen, NADPH, H4BP, ( 09480167 to T.N., and 8780599 and 11780455 to T.I.), and by the and Ca2+-calmodulin (1, 6, 14). The first step is the forma National Institutes of Health (GM 42025 to R.A.S.). The XAS data tion of NƒÖ-hydroxy-L-arginine (NOHA) as an enzyme-bound were collected at SSRL. which is operated by the Department of En ergy. Division of Chemical Sciences. The SSRL Biotechnology pro intermediate (1, 14, 27). It has been proposed, by analogy to gram is supported by the National Institute of Health. Biomedical P450 mechanisms, that the conversion of L-arginine to Technology Program. Division of Research Resources. NOHA requires reduction of the ferriheme, binding of O2 to 2 To whom correspondence should be addressed: Robert A. Scott, form the ferrous-O2 adduct, and activation of O2 to form a Phone +1-706-542-2726, Fax: +1-706-542-9454, E-mail: rscott@ reactive high-valent oxo-ferryl porphyrin, a ƒÎ-cation radical uga.edu; Toshio Iwasaki, Phone: +81-3-3822-2131 (Ext. 5216), Fax; species which allows oxygen insertion into L-arginine (1, 14, +81-3-5685-3054, E-mail: iwasaki/[email protected] 28). The second step, i.e. conversion of NOHA to citrulline Abbreviations: eNOS, endothelial nitric oxide synthase; EPIC elec tron paramagnetic resonance; EXAFS, extended X-ray absorption and NO, is thought to require the binding of O2 to the fer fine structure; FT, Fourier transform; H4BP, (6R)-5,6.7.8-tetrahydro rous enzyme, and the involvement of the ferric peroxide -L- biopterin; iNOS, inducible nitric oxide synthase; NO. nitric oxide; intermediate and an NO-L-arginine radical to form a puta NOHA, NƒÖ-hydroxy-L-arginine; NOS, nitric oxide synthase; nNOS, tive iron-peroxo-NOHA radical intermediate (1, 14, 28). neuronal nitric oxide synthase; SSRL, Stanford Synchrotron Radia Although a possible two-step oxygenation mechanism for tion Laboratory; XAS, X-ray absorption spectroscopy. NOS has been proposed on the basis of analogy to P450-N-

c 2001 by The Japanese Biochemical Society hydroxylation and P450-aromatase systems, there are sub-

Vol. 130, No. 2, 2001 191 192 N.J. Cosper et al.

tie differences between NOS isoforms and of analytical grade. P450, which suggest that the NOS reaction may be more A heterologous expression system for the full-length complex. First, there is little homology in wild-type nNOS in Escherichia coli was constructed w th sequence or overall protein folding between the NOS oxyge the pCWori+ vector (47-49) and the chaperonin expression nase domain and cytochrome P450, thereby excluding NOS vector pKY206 (pACYC184; Nippon Gene, Toyama) carry- from the cytochrome P450 gene family All NOS isoforms ing the E. coli chaperonin groELS genes (kindly provided are each catalytically active only as a homodimer (1, 6, 8, by Dr. K. Ito, Kyoto University), as reported previously (31). 29), and mutagenesis studies have indicated that appropri The recombinant nNOS produced in E. coli strain BL21 ate cross-talk between the N-terminal hook region and the was purified on ice or at 4•Ž essentially as described in the substrate-binding distal pocket of NOS isoforms is essential literature (49), except that purification was conducted by for proper activation and regulation of the citrulline- and 2•Œ,5•Œ-ADP Sepharose 4B column chromatography (Amer NO-forming activity (12, 30-32). Second, all NOS isoforms sham Pharmacia Biotech), followed by Sephacryl S300HR show an absolute requirement of H4BP for NO synthase and DEAF Sepharose Fast Flow column chromatography activity (33, 34). The exact role of H4BP in NOS catalysis (Amersham Pharmacia Biotech) (31), and that the over- remains controversial. Recent studies have suggested a night dialysis step was omitted. H4BP (30ƒÊM) and L-argi possible redox function for the H4BP (13, 35, 36), nine (1mM) were supplied during the purification. The but the redox cycling of H,BP in the NOS catalytic cycle catalytic activity and purity of the purified enzyme were seems to be different from that observed with the aromatic comparable to those previously reported for recombinant amino acid hydroxylase reaction, which requires a non nNOS by others (49). heme iron center in close proximity to H4BP (13, 36). NOS activity was measured by monitoring the conver To fully understand the mechanismof NO synthesis,it is sion of [14C-U]-L-Arg to [14C-U]-L-citrulline, as described essential to determine the changes that the catalyticcenter previously (50). The standard assay was performed at 25•Ž undergoes during the reaction cycle(28, 37). The iron coor in an assay mixture containing 16.7mM HEPES-NaOH dination structures of a variety of heme proteins including buffer, pH 7.4, 4.2mM Tris-HCl buffer, pH 7.4, 667ƒÊM cytochromesP450 and chloroperoxidasehave been studied EDTA, 167ƒÊM EGTA, 667 ƒÊM DTT, 16.7ƒÊM [14C-U]-L- by X-ray absorption spectroscopy(XAS) (37-40). The metal Arg, 667ƒÊM NADPH, 1.2mM CaCl, 6.7ƒÊg of calmodulin, ligand bond distancesobtained by XASand high-resolution 1.25ƒÊM FAD, 1.25ƒÊM FMN, 2.5ƒÊM H,BP and the en X-ray crystallographicanalysis are in close agreement for zyme, in a total volume of 30ƒÊl. The specific activity of [14C-U]-L-Arg several different oxidation and spin states of cytochrome -Are used in the assays was 11.84GBq/mmol. P450 (37). Extensive XASanalyses have been conducted Absorption spectra were recorded with a Hitachi U3210 with cytochromesP450 and chloroperoxidaseto elucidate spectrophotometer or a Beckman DU-7400 spectrophotom the change in the iron coordinationin relation to the cata eter. Electron paramagnetic resonance (EPR) spectra of lytic mechanism and to elucidate the structures of stable several different batches of purified nNOS were measured intermediate species (37-42).Although several X-ray crys at JEOL, Tokyo, with a JEOL JES-TE200 spectrometer tal structures of eNOS and iNOS oxygenatedomains com equipped with an ES-CT470 Heli-Tran cryostat system, in plexed with substrate and inhibitors have been reported which the temperature was monitored with a Scientific recently (10-13, 43), the NOS holoprotein is a complex, Instruments digital temperature indicator/controller Model multidomain flavohemoproteinwhose flavin and heme cen 9650 and the magnetic field was monitored with a JEOL ters are probably positioned near each other (21). No de NMR field meter ES-FC5. EPR spectroscopic measure tailed structural information on the heme site has been ments were also carried out with a JEOL JEX-RE1X spec obtained for full-length NOS isoforms,and no XAS studies trometer equipped with an Air Products model LTR-3 Heli on any oxidation or spin state of any NOS isoform have Tran cryostat system, in which the temperature was moni been reported so far. As an initial step toward understand tored with a Scientific Instruments series 5500 tempera ing the structural changesof the heme site during the cata ture indicator/controller, as reported previously (51). All lysis of NOS isoforms,we report herein XAS analysis of EPR data were processed with KaleidaGraph software ver. the ferriheme site of substrate-boundfull-length nNOS. We 3.05(Abelbeck Software). also examine the heme site structure of nNOS upon bind XAS data were collected at the Stanford Synchrotron ing a divalent cation inhibitor, Cu2+,that strongly blocks Radiation Laboratory (SSRL) with the SPEAR storage ring the NOS activity of constitutiveNOS isoforms(44-46), and operating in a dedicated mode at 3.0 GeV (Table I) . The may have an impact on controllingNOS and guanylylcycl iron concentrations in the samples for XAS measurement ase activity in mammalian cells. were ~1mM. The XAS samples were prepared as outlined for bacterial cytochrome aa3 (52) with the following modifi EXPERIMENTAL PROCEDURES cations. Purified nNOS preincubated with 10mM substrate (L-arginine or NOHA) and 1mM H4BP on ice in either the Calmodulin, FAD, FMN, L-Arg, and L-citrulline were from presence or absence of CuCl2 was immediately precipitated with 40mM potassium phosphate buffer Sigma Chemicals, and 6(R)-5,6,7,8-tetrahydro-L-biopterin , pH 7.4, contain (H4BP) was obtained from the Schircks Laboratories (Jona, ing ammonium sulfate (30% saturation) , 10mM substrate Switzerland). [14C-U]-L-Arg was obtained from Dupont, (L-arginine or NOHA), and 1mM H4BP at 4•Ž. The precipi New England Nuclear. 2•Œ,5•Œ-ADP Sepharose 4B, Sephacryl tate was rapidly rinsed with 20mM potassium phosphate S-300HR, and DEAF Sepharose Fast Flow were purchased buffer, pH 7.4, containing 10mM substrate (L-arginine or from Pharmacia Biotech, and Ampure SA from Amersham NOHA) and 1mM H4BP. Further concentration was Life Science. Water was purified with a Milli-Q purification achieved by placing the samples on ice under a stream of system (Millipore). Other chemicals used in this study were dry nitrogen gas. The entire purification and concentration

J. Biochem. XASAnalysis of the Ferriheme Site in nNOS 193

processes were performed in 24 h. The resulting samples (-1mM iron), containing -30% (v/v) glycerol , were transf RESULTS AND DISCUSSION erred directly to 24 •~ 3 •~ 1mm polycarbonate cuvettes with a Mylar-tape front window and frozen in liquid nitro EPR Spectra-The resting recombinant mouse nNOS gen. The spin state of the nNOS ferriheme used for XAS anal purified in the presence of H4BP and L-arginine showed cit rulline-forming activity of -220-280nmol/mg/min at 25•Ž ysis was checked by EPR spectroscopy. Because holo nNOS with an apparent Km for L-arginine of 1.3ƒÊM (data not is a complex multidomain flavo-metalloenzyme , one of the shown). A broad Soret band at around 395 run in the visible most difficult steps in the sample preparation is concentra absorption spectrum was observed for nNOS in the pres tion to levels suitable for XAS analysis. Extensive concen ence of 0.4mM L-arginine and 30ƒÊM H4BP (data not tration with pressure filtration for extended periods of time shown) (31). The EPR spectrum of the ferriheme center in often gave nNOS samples with a significantly higher con the resulting enzyme confirms the high-spin species, char tent of high-spin ferric P420 form at g = 6.02 due to disrup acterized by apparent g values (gx = 7.59, gy = 4.08, and g2 = tion of the Fe-S(Cys) bond (31, 53-55) and/or low-spin 1.81) and rhombicity, defined as the ratio of the rhombic ferriheme species at g = 2.45-2.41, 2.28 , and 1.91-1.90, and axial zero field splitting parameters, E/D) (E/D = likely due to incomplete water coordination as the axial 0.073) (Fig. IA). sixth ligand, as in the case of the concentrated resting Preincubation of the resting enzyme with the intermedi enzyme (typically -30-50%; data not shown). Some of these ate (NOHA) and H4BP resulted in conversion of the ferri samples exhibited significant spin-state heterogeneity with heme species to another different high-spin form with g- respect to the ferriheme site, and thus were unsuitable for values gx = 7.70, gy = 3.99, and g2 = 1.80, and greater rhom XAS analysis (data not shown). On the other hand, ammo bicity (E/D = 0.077; Fig. 1C). The visible absorption spec nium sulfate precipitation in the presence of the substrate trum of the NOHA-bound form is indistinguishable from and H4BP gave samples with predominantly high-spin fer that of the arginine-bound form, displaying a broad Soret riheme species (>90%) exhibiting characteristic EPR spec band at around 395nm (data not shown). Thus, the bind tra (Fig. 1) that were stable during sample manipulation ing of either L-arginine or NOHA causes an EPR-detectable and irradiation at SSRL. local conformational change in the enzyme that produces EXAFS analysis was performed with the EXAFSPAK ligand-specific high-spin species (20). The narrow line- software (courtesy of G.N. George; www-ssrl.slac.stanford. widths of the EPR spectra suggest that the substrate target edu/exafspak.html) according to standard procedures (56). atom in high-spin ferric nNOS is located at a well-ordered Curve-fitting analysis was performed as described previ position near to the ferriheme iron. The geometry and ously by others (57, 58). Briefly, the crystallographic coordi nates for 2,3,7,8,12,13,17,18-octaethylporphyrinato-iron(II) (59) were imported into Chem 3D Pro (Cambridge Scientific) and edited to include only one fourth of porphyrin, excluding hydrogen atoms. The coordinates for the remaining atoms were imported into FEFF v7.02 (60) to calculate EXAFS phase and amplitude functions for both single- and multiple- scattering paths. These functions were then incorporated into EXAFSPAK curve-fitting software to fit the data.

TABLE I. X-ray absorption spectroscopic data collection.

Fig. 1. EPR spectra at 7 K of L-arginine-bound (A) and NOHA- bound nNOS (C), and the XAS samples of L-arginine-bound (B) and NOHA-bound nNOS (D). All nNOS samples used in this work were purified in the presence of H4BP cofactor. The minor fer riheme species at g = 6.02 (less than 2-3% of the total ferriheme) was formed during concentration of the nNOS samples. A signal at g = 4.3 from high-spin rhombic Fe3+ represents a small amount of ad ventitiously bound non-heme iron. The loss of a radical feature at g = 2.0 for the XAS samples is due to the long-term storage of the sam

ples at cryogenic temperature (below -80•Ž). Instrument settings: a The 13-element Ge solid-state X-ray fluorescence detector at microwave power, 1mW; modulation amplitude, 1mT; variable SSRL is provided by the NIH Biotechnology Research Resource. gain; the g values are indicated in the figure.

Vol. 130, No. 2, 2001 194 N.J. Cosper et al.

structures of these high-spin ferriheme species were fur several divalent cations, such as Cu2+, Pb2+, and Zn2+ (44). ther investigated by X-ray absorption spectroscopy. Cu2+ uptake by glial cells selectively inhibits nNOS activity, Fe K-Edge XAS Analysis-The similarity between the Fe and results in a decrease of the basal intracellular cGMP K-edge X-ray absorption spectra of the NOHA- and argin levels and an increase of iNOS-mediated NO synthesis ine-bound forms (Fig. 2A) implies a general structural and through transcriptional activation of iNOS mRNA (61). electronic equivalence between the iron coordination sites Cu2+ also suppresses the eNOS activity extracted from pul of the two forms. The Fe EXAFS for L-arginine-bound monary arterial endothelial cells (45), but in the presence nNOS are best fit assuming a sulfur ligand at 2.30 A and of extracellular Ca2+, it relaxes vessels and results in the four nitrogen atoms (from the plane of porphyrin) at 2.03 A accumulation of eNOS-mediated cGMP in the pulmonary (Fit 1; Table II and Fig. 3A), while the EXAFS for NOHA- artery employing extrapulmonary arterial rings (62). The bound nNOS are best fit assuming a sulfur atom at 2.29 A Cu2+-mediated increase of NO production via eNOS activa and four porphyrin nitrogen atoms at 2.02 A (Fit 2; Table II tion is due to Ca2+ influx into the pulmonary arterial endot and Fig. 3B). helial cells (45, 63). Thus, Cu2+ has been proposed to play Fe K-edge EXAFS of the high-spin ferric states of cyto regulatory roles in controlling NOS and guanylyl cyclase chrome P450camand chloroperoxidase shows the presence of activity in mammalian cells. a sulfur donor ligand coordinated to the iron at 2.23-2.26 We confirmed that several divalent cations, such as the and 2.30 A, respectively (37-40). The observed Fe-S bond Cu2+, Mn2+, and Zn2+ ions, at 1mM completely inhibit distances (2.29 A) of full-length nNOS and chloroperoxi nNOS activity (data not shown). It has been reported that dase are longer than that of high-spin ferric cytochrome the Zn2+ ion perturbs the environment of the heme iron in P450cam,but shorter than those of known Fe(III)-S(thiolate) nNOS (64), and Perry and Marietta (46) have postulated heme iron model systems (37-39). that exogenous Fe2+ might bind in close proximity to the Effect of a Divalent Cation Inhibitor on the Iron Coordi heme and/or H,BP center of NOS. We therefore examined nation Environment of nNOS-It has been reported that the effect of Cu2+ on the immediate heme coordination envi constitutive NOS activity in rat cerebellum is inhibited by ronment of nNOS by X-band EPR, and Cu and Fe K-edge XAS. The XAS technique "sees" the average coordination environment of all atoms of a given element, and is suitable for analyzing any structural changes of the nNOS metal centers upon Cu2+ binding. The effect of exogenous Zn2+ or Fe2+ was not investigated in this study because of the pres ence of a tetrahedral tetrathiolate zinc center at the dieter interface and heme iron in NOS oxygenase domains (10, 11, 13, 43, 65). The Cu2+-inhibited form of arginine/H4BP-bound nNOS yields high-spin ferriheme EPR signals that are similar to those observed for arginine/H4BP-bound nNOS (Fig. 4). An additional broad Cu2+ signal is also observed in the g = 2 region. The difference between the line positions of the high-spin ferriheme in the absence and presence of Cu2+ is less than the line widths. The presence of the low-spin fer riheme signals at g = 2.41, 2.28, and 1.91 for the Cu2+- inhibited forms indicates that a small amount of low-spin ferriheme is formed in the presence of excess Cu2+ ion (Fig. 5), as was previously reported for inhibition of nNOS by Zn2+ (64). However, the observable spin state conversion of Fig. 2. Fe K-edge X-ray absorption spectra for L-arginine- the nNOS ferriheme with the Cu2+ ion was minimal (esti bound nNOS ferriheme (A, B, C; solid) compared with NOHA- bound nNOS (A; dashed), L-arginine-bound nNOS containing mated to be approximately 15% of the total ferriheme) even 1 eq. Cu2+ (per heme) (B; dashed), or L-arginine-bound nNOS in the presence of a saturating amount of Cu2+, and the incubated with excess (ca. 2 eq.) Cu2+ (C; dashed). degree of the spin state conversion is not consistent with

TABLE II. Curve fitting results for Fe EXAFS.a

Ras is the metal-scatterer distance. ƒÐ2as is a mean square deviation in Ras, af is a normalized error (chi-squared):

J. Biochem. XAS Analysis of the Ferriheme Site in nNOS 195

the total loss of the citrulline formation activity. bond distance of 2.30 A, indicating no significant stretching Fe K-edge XAS indicates that the addition of excess Cull of the Fe-S (thiolate) bond between the two high-spin ferric only slightly affects the immediate ferriheme coordination forms (Fig. 6). The substrate is not a ligand to the high- environment (Fig. 2, B and C). There is no evidence of Fe- spin, five-coordinate ferriheme iron of full-length nNOS, Cu (Fig. 3) or Cu-Fe scattering in the EXAFS and FT spec consistent with inferences based on the results obtained tra. Furthermore, Cu K-edge absorption spectra and with other spectroscopic techniques (17-22, 24-26). These EXAFS (data not shown) indicate that nNOS-bound Cull structural features are in good agreement with the previ ions are indistinguishable from spectra recorded in the ously reported crystal structures of the arginine-bound presence of excess Cu2+, and no evidence for the hyperfine eNOS and iNOS oxygenase domains in the presence of coupling of Cu2+ with 14N nuclei was obtained by EPR (Figs. H4BP [showing the presence of a cysteinate ligand with a 4 and 5). Thus, the Cu2+ inhibition slightly affects the spin state equilibrium but causes no significant change in the immediate ferriheme coordination environment of nNOS. Contrary to the earlier proposals by others (46, 64), these results suggest that the Cu2+ responsible for the inhibitory effect binds away from the heme site to cause a slight con formational change and to block the citrulline formation activity. NOS isoforms are compex multidomain proteins with multiple redox centers, whose overall conformation is high ly sensitive to binding of substrates and cofactors such as H4BP, which binds in close proximity to the heme site (10- 13,43). Because XAS indicates that the divalent metal does not bind near the heme, our results provide no support for the aromatic amino acid hydroxylase reaction of H4BP in nNOS catalysis, which would require a non-heme iron cen ter in close proximity to the H4BP center. This is in line with other indirect evidence inferred from recent studies (43, 66, 67). Structure of the High-Spin Ferriheme Site in Substrate- bound nNOS-Central to all proposed catalytic mecha nisms for the two-step oxygenase reaction by NOS isoforms is the close proximity of the reactive guanidino nitrogen of the substrate to the heme iron. The present XAS analysis Fig. 4. EPR spectra at 6.2 K of the high-spin ferriheme center demonstrates that the high-spin ferriheme site geometry of trarginine bound nNOS in the presence of H,BP (A), and and Fe-S(thiolate) bond distances of the arginine-bound in the presence of 1 eq. Cu2+ (B) or excess (ca 2 eq.) Cu2+ (C). and NOHA-bound forms of full-length nNOS in the pres Instrument settings: microwave power, I mW; modulation ampli ence of H4BP are essentially identical, with a Fe-S(thiolate) tude, 1mT; the g values are indicated in the figure.

Fig. 3. X-ray absorption fine struc ture (left) and Fourier transforms (right, over k=2-11.5 A-1) for L- arginine-bound nNOS ferriheme (A, solid) compared with the cal culated spectra for Fit 1, Table II (A; dashed), NOHA-bound nNOS (B, solid) compared with the cal culated spectra for Fit 2, Table H (B; dashed), L-arginine-bound nNOS containing 1 eq. Cu2+ (per total heme) (C; solid) compared with the calculated spectra for Fit 3, Table II (C; dashed), or L- arginine-bound nNOS containing excess (ca 2 eq.) Cu2+ (D; solid) compared with the calculated spectra for Fit 4, Table II (D; dashed).

Vol. 130, No. 2, 2001 196 N.J. Cosper et al.

Fig, 6. Schematic illustration of the heme site geometry of L- arginine (A) and NOHA (B) bound nNOS ferriheme. This illus tration is based on the structural information on the high-spin ferri heme coordination environment of the ligand-bound full-length nNOS obtained through the present XAS analysis, and the Q-band pulsed 15N ENDOR analysis by Tierney et al. (69, 70).

droxylated nitrogen of NOHA relative to the high-spin fer riheme iron of full-length nNOS has been determined by Fig. 5. EPR spectra at 20 K of L-arginine bound nNOS in the very recent Q-band pulsed 15N ENDOR analysis, which presence of H4BP (E), and in the presence of 1 eq. Cu2+ (A) or showed that "N of singly labeled NOHA at the hydroxy excess (ca. 2 eq.) Cu2+(B). Subtraction of spectrum A from B lated nitrogen is 3.8 A from the ferriheme iron, i.e. closer gives an EPR spectrum attributed to adventitiously bound Cu2+ (C), and subtraction of resulting spectrum C fromA or B than the corresponding guanidino 15NƒÖ of L-arginine (4.05 gives an identical EPR spectrum attributed to the minor A) (69, 70). In conjunction with the common structure and low-spin ferriheme species of nNOS (D). The EPR samples used geometry of the high-spin ferriheme site of the substrate- are the same as those in Fig. 2. Instrument settings: microwave bound nNOS revealed in this study, it can be concluded power, 1mW, modulation amplitude, 1mT; the g values are indi that the high-spin ferric nNOS holds the substrate target cated in the figure. atom in a well-ordered position near the heme iron (Fig. 6), which may be favorably located for subsequent hydroxyla tion by an iron-bound oxygenic species. Further XAS analy Fe-S distance of about 2.3 A (11-13, 43, 68)], and the iNOS sis with different oxidation and spin-states of nNOS, along oxygenase domain complexed with NOHA (reported during with other analytical techniques, will firmly establish the the preparation of this paper (68)), and therefore likely detailed structural changes of the heme site geometry and common among the NOS isoforms. In the 2.4-A resolution the reaction intermediates in the catalytic events of NOS crystal structure of the human eNOS oxygenase domain, isoforms. the iron was found to be 0.65 A out of the plane of the four

pyrrole nitrogen toward the cysteinate ligand (11). Inter We wish to thank Drs. K. Tamura and T. lizuka (The Institute of estingly, the Fe-S (thiolate) bond length of full-length nNOS Physical and Chemical Research) for allowing us to utilize the X- is similar to that of the high-spin ferric state of chloroperox band EPR facility, Dr T. Oshima (Tokyo University of Pharmacy idase, but longer than that of cytochrome P450cam(37-39). and Life Science) for allowing us access to the large-scale fermen This seems to be in line with the results of previous reso ter facility, and Y. Hayashi (Nippon Medical School) for the excel nance Raman analysis of the ferrous CO-bound forms of lent technical assistance. We also thank Marion Cooper for the substrate-bound nNOS by Wang et al. (24), suggesting that help in acquiring some of the XAS data. the bonding of the cysteinate to the heme iron may be weaker, as observed in chloroperoxidase, than in cyto REFERENCES chrome P450cam, on the basis of the inverse correlation 1. Griffith, O. and Stuehr, D.J. (1995) Nitric oxide synthases: between the Fe-CO and C-O stretching modes. properties and catalytic mechanism. Annu. Rev. Physiol. 57, Finally, the absence of XAS-detectable differences in the 707-736 heme site structure and geometry or the conformation of 2. Ignarro, L.J. (1996) Physiology and pathophysiology of nitric the enzyme-bound substrate between the L-arginine- and oxide. Kidney lot. 55, S2-S5 NOHA-bound forms indicates that the position of the reac 3. Marietta, M.A. (1994) Nitric oxide synthase: aspects concerning structure and catalysis. Cell 78 tive guanidine nitrogen (NƒÖ) of L-arginine and NOHA in , 927-930 4. Moncada, S. and Higgs, E.A. (1995) Molecular mechanisms and the binding site, and the resulting orientation of the fer therapeutic strategies related to nitric oxide. FASEB J. 9. rous-O2, adduct, rather than the heme site structure and 1319-1330 geometry, specify the two-step hydroxylation reactions in 5. Nathan, C. and Xie, Q.-w. (1994) Regulation of biosynthesis of the catalytic site of full-length nNOS. The position of the nitric oxide. J. Biol. Chem. 269, 13725-13725 reactive guanidino nitrogen (NƒÖ) of L-arginine and the by- 6. Stuehr, D.J. (1999) Mammalian nitric oxide synthases . Bio-

J. Biochem. XAS Analysis of the Ferrihenre Site in nNOS 197

chim. Biophys.Acta 1411, 217-230 10987-10992 7. Stuehr, D.J. (1997) Structure-function aspects in the nitric 26. Fan, B., Wang, J., Stuehr, D., and Rousseau, D.L. (1997) NO oxidesynthases. Annu. Rev.Pharmacol, 37,339-359 synthase isozymes have distinct substrate binding sites. Bio 8. RMasters, B.S.S., McMillan, K., Sheta, E.A., Nishimura, J.S., chemistry 36.12660-12665 oman,L.J., and Martasek, P. (1996)Neuronal nitric oxide syn thase, a modular enzyme formed by convergent evolution: 27. Iwanaga, T., Yamazaki, T., and Kominami, S. (1999) Kinetic studies on the successive reaction of neuronal nitric oxide s structure studies of a cysteine thiolate-ligandedheme protein ynthase from L-arginine to nitric oxide and L-citrulline. Biochem that hydroxylates L-arginineto produce NO as a cellular sig istry 38, 16629-16635 nal. FASEBJ. 10, 552-558 28. Sono, M., Roach, M.P., Coulter. E.D., and Dawson, J.H. (1996) 9. Knowles.R.G. and Moncada,S. (1994)Nitric oxidesynthases in Heme-containing oxygenases. Chem. Rev. 96, 2841-2887 mammals.Biochem. J. 298, 249-258 29. Klatt, P., Schmidt, K., Lehner, D., Glatter, O., Bachinger, H.P., 10. Crane, B.R., Rosenfeld,R.J., Arvai, A.S.,Ghosh , D.K.,Ghosh, and Mayer, B. (1995) Structural analysis of porcine brain nitric S., Tainer J.A., Stuehr, D.J., and Getzoff.E.D. (1999)N-termi oxide synthase reveals a role for and L- nal domain swapping and metal ion bindingin nitric oxidesyn arginine in the formation of an SDS-resistant diner. EMBO J. thase dimerization.EMBO J. 18, 6271-6281 14,3687-3695 11. Fischmann, T.O., Hruza, A., Niu, X.D., Fossetta, J.D., Lunn, 30. Rodriguez-Crespo, I., Moenne-Loccoz, P., Loehr T.M., and Ortiz C.A., Dolphin, E., Prongay,A.J., Reichert, P., Lundell, D.J., de Montellano, P.R. (1997) Endothelial nitric oxide synthase: Narula, S.K., and Weber,P.C. (1999)Structural characteriza modulations of the distal heave site produced by progressive N- tion of nitric oxide synthase isoformsreveals striking active terminal deletions. Biochemistry 36, 8530-8538 site conservation.Nat. Struct. Biol. 6, 233-242 31, Iwasaki, T., Hon, H., Hayashi, Y., and Nishino, T. (1999) Modu 12. Crane, B.R., Arvai, A.S., Ghosh, D.K., Wu, C., Getzoff F.D., lation of the remote hence site geometry of recombinant mouse Stuehr, D.J., and Tamer, J.A. (1998) Structure of nitric oxide neuronal nitric oxide synthase by the N-terminal hook region. synthase oxygenase diver with ptein and substrate. Science J. Biol. Chem. 274, 7705-7713 279.2121-2126 32, Ghosh, D.K., Crane, B.R., Ghosh, S., Wolan, D., Gachhui, R., 13. Roman, C.S.,Li, H., Martasek, P., Kral,V., Masters, B.S.S.,and Crooks, C., Presta, A.. Tamer, J.A., Getzoff, E.D., and Stuehr, Poulos.T.L. (1998) Crystal structure of constitutive endothelial D.J. (1999) Inducible nitric oxide synthase: role of the N-termi nitric oxidesynthase; a paradigm for pterin functioninvolving nal ƒÀ-hairpin hook and pterin-binding segment in dimerization a novelmetal center Cell 95, 939-950 and tetrahydrobiopterin interaction. EMBO J. 18, 6260-6270 14. Marietta, M.A. (1993) Nitric oxide synthase structure and 33. Kwon, N.S., Nathan, C.F., and Stuchr, D.J. (1989) Reduced mechanism.-J.Biol. Chem.268, 12231-12234 biopterin as a cofactor in the generation of nitrogen oxides by 15. Abu-Soud,H.M., Gachhui, R., Raushel, F.M.,and Stuehr, D.J. marine macrophages. J. Biol. Chem. 264. 20496-20501 (1997)The ferrous-dioxycomplex of neuronal nitric oxide syn 34. Stuehr. D.J., Kwon, N.S., Nathan, C.F. Griffith, O.W., Feldman, thase: divergent effects of L-arginine and tetrahydrobiopterin P.L., and Wiseman, J. (1991) NƒÖ-hydroxy-L-arginine is an inter on its stability J. Biol. Chem.272, 17349-173531 mediate in the biosynthesis of nitric oxide from L-arginine. J. 6. Ledbetter A.P., McMillan, K., Roman, L.J., Masters, B.S.S., Biol. Chem. 266,6259-6263 Dawson.J.H., and Sono,M. (1999) Low-temperaturestabiliza 35. Witteveen, C.F.B., Giovanelli, J., and Kaufman, S. (1999) Reac tion and spectroscopiccharacterization of the dioxygencomplex tivity of tetrahydrobiopterin bound to nitric-oxide synthase J. of the ferrous neuronal nitric oxide synthase oxygenasedo Biol. Chem. 274,29755-29762 main. Biochemistry38, 8014-8021 36. Hurshman, A.R., Krebs, C., Edmondson, D.E., Huynh, B.H., 17. Stuehr, D.J. and Ikeda-Saito,M. (1992) Spectral characteriza and Marietta, M.A. (1999) Formation of a pterin radical in the tion of brain and macrophage nitric oxide synthases:cyto reaction of the home domain of inducible nitric oxide synthase chrome P-450-like hemoproteinsthat contain a flavin semi with oxygen. Biochemistry 38, 15689-15696 quinoneradical J. Biol. Chem.267, 20547-20550 37. Dawson, J.H. and Sono. M. (1987) Cytochrome P-450 and chlo 18. Wang. J., Stuehr, D.J., Ikeda-Saito, M., and Rousseau, D.L. roperoxidase: thiolate-ligated heme enzymes: spectroscopic de ( 1993)Home coordination and structure of the catalytic site in termination of their active site structures and mechanistic im nitric oxidesynthase. J. Biol. Chem.268, 22255-22258 plications of thiolate ligation. Chem. Rev. 87, 1255-1276 19. Soon. M., Stuehr, D.J., Ikeda-Saito. M., and Dawson, J.H. 38. Cramer. S.P., Dawson, J.H., Hodgson, K.O., and Hager, L.P. (1995) Identification of nitric oxide synthase as a thiolate (1978) Studies on the ferric forms of cytochrome P-450 and ligand heme protein using magnetic circular dichroism spec chloroperoxidase by extended X-ray absorption fine structure: troscopy:comparison with cytochromeP-450-cam and chlorop characterization of the Fe-N and Fe-S distances. J. Am. Chem. eroxidase.J. Biol. Chem.270, 19943-19948 Soc. 100, 7282-7290 20. Salerno.J.C., McMillan,K., and Masters, B.S.S. (1996)Binding 39. Hahn, J.E., Hodgson, E.G., Andersson, L.A., and Dawson, J.H. of intermediate, product, and substrate analogs to neuronal (1982) Endogenous cysteine ligation in ferric and ferrous cyt nitric oxide synthase ferriheme is sensitive to ligand-specific ochrome P-450: direct evidence from X-ray absorption spectros effectsin the L-argininebinding site. Biochemistry35, 11839- copy J. Biol. Chem. 257,10934-10941 11845 40. Obayashi, E., Tsukamoto, K., Adachi, S.-i., Takahashi, S., 21. Galli, C., MacArthur, R., Abu-Soud,H.M., Clark, P., Steuhr, Nonnura, M., lizuka, T., Shoun, H., and Shiro, Y. (1997) Unique D.J.,and Brudvig. G.W. (1996)EPR spectroscopiccharacteriza binding of nitric oxide to ferric nitric oxide reductase from tion of neuronal NO synthase.Biochemistry 35, 2804-2810 Fusarium oxysporum elucidated with infrared, resonance 22. Tsai, A.-L.,Berka, V., Chen, P.-F.,and Palmer, G. (1996)Char Roman, and X-ray absorption spectroscopies. J. Am. Chem. Soc. acterization of endothelial nitric-oxide synthase and its reac 119,7807-7816 tion with ligand by electron paramagnetic resonance spectro 41. Shiro, Y., Sato, F., Suzuki, T., lizuka, T., Matsushita, T., and scopy,J. Biol. Chem.271,32563-32571 Oyanagi, H. (1990) X-ray absorption spectral study of ferric 23. Poulos,T.L. (1996) The role of the proximalligand in heme en (nigh-spin : XANES evidences for coordination zymes.J. Biol.lnorg. Chem. 1. 356-359 structure of the here iron. J. Am, Chem, Soc. 112, 2921-2924 24. Wang, J.,Stuehr, D.J.,and Rousseau.D.L. (1997) Interactions 42. Liu. H.I., Soon, M., Kadkhodayan, S., Hager, L.P., Hedman, B., betweensubstrate analogues and home ligands in nitric oxide Hodgson, K.O., and Dawson, J.H. (1995) X-ray absorption near synthase.Biochemistry 36. 4595-4606 edge studies of cytochrome P-450-CAM, chloroperoxidase, and 25. Migita, C.T., Salerno, J.C., Masters, B.S.S., Martasek. P., myoglobin: direct evidence for the electron releasing character McMillan,K., and Ikeda-Saito, M. (1997) Substrate binding of a cysteine thiolate proximal ligand. J. Biol. Chem. 270, inducedchanges in the EPR spectra of the ferrous nitric oxide 10544-10550 complexesof neuronal nitric oxide synthase. Biochemisty 36. 43. Li, H., Raman, C.S., Closer, C.B., Blasko, E., Young, T.A., Par-

Vol. 130, No . 2, 2001 198 N.J. Cosper et al.

kinson, J.F., Whitlow, M., and Poulos, T.L. (1999) Crystal struc heart Met- and deoxymyoglobin using multiple-scattering tures of zinc free and -bound heme domain of human inducible XAFS analyses. Inorg. Chem. 37, 5743-5753 nitric-oxide synthase. J. Biol. Chem. 274, 21276-21284 58. Rich, A.M., Armstrong, R.S., Ellis, P.J., and Lay, P.A. (1998) 44. Quinn, M.R. and Harris, C.L. (1995) Lead inhibits Ca2+-stimu Determination of the Fe-ligand bond lengths and Fe-N-O bond lated nitric oxide synthase activity from rat cerebellum . Neuro angles in horse heart ferric and ferrous nitrosylmyoglobin sci. Lett. 196, 65-68 using multiple-scattering XAFS analysis. J. Am. Chem. Soc. 45. Demura, Y., Ameshima, S., Ishizaki, T., Okanuua, S., Miyamori, 120,10827-10836 I., and Matsukawa, S. (1998) The activation of eNOS by copper 59. Strauss, S.H., Silver, M.E., Long, K.M., Thompson, R.G., Hud ion (Cu2+) in human pulmonary arterial endothelial cells gens, R.A., Spartalian, K., and fliers, J.A. (1985) Comparison of (HPAEC) Free Radic. Biol. Med. 25, 314-320 the molecular and electronic-structures (2,3,7,8,12,13,17,18- 46. Perry, J.M. and Marietta, M.A. (1998) Effects of transition met octaethylporphyrinato)iron(II) and (trans-7,8-dihydro-2,3,7,8, als on nitric oxide synthase catalysis. Proc. Natl. Acad. Sci. 12,13,17,18-octaethylporphyrinato)iron(II). J. Am. Chem. Soc. U.S.A. 95. 11101-11106 107, 4207 47. Muchmore, D.C., McIntosh, L.P., Russell, C.B., Anderson, D.E., 60. Zabinsky, S.I., Rehr, J.J., Ankudinov, A., Albers, R.C., and Eller, and Dahlquist, F.W. (1989) Expression and nitrogen-15 labeling M.J. (1995) Multiple-scattering calculations of X-ray absorption of proteins for proton and nitrogen-15 nuclear magnetic reso spectra. Phys. Rev. B52,2995-3009 nance. Methods Enzymol. 177. 44-73 61. Colasanti, M., Persichini, T., Venturini, G., Polticelli, F., and 48. Gerber, N.C., and Ortide Montellano, P.R. (1995) Neuronal Musci, G. (2000) Modulation of the nitric oxide pathway by cop nitric oxide synthase: expression in Escherichia coli, irrevers per in glial cells. Biochem. Biophys. Res. Commun. 275, 776- ible inhibition by phenyldiazene, and active site topology J. 782 Biol. Chem. 270, 17791-17796 62. Ohnishi, T., Ishizaki, T., Sasaki, F., Ameshima, S., Nakai, T., 49. Roman. L.J., Sheta, E.A., Martasek, P., Gross, S.S., Liu, Q., and Miyabo, S., and Matsukawa, S. (1997) The effect of Cu2+on rat Masters, B.S.S. (1995) High-level expression of functional rat pulmonary arterial rings. Eur J. Pharmacol. 319,49-55 neuronal nitric oxide synthase in Escherichia coli. Proc. Natl. 63. Demura, Y., Ishizaki, T., Ameshima, S., Okamura, S., Hayashi, Acad. Sci. USA 92,8428-8432 T., Matsukawa, S., and Miyamori, I. (1998) The activation of 50. Hori, H., Iwasaki, T., Kurahashi, Y., and Nishino, T. (1997) Cal nitric oxide synthase by copper ion is mediated by intracellular cium-dependent inactivation of neuronal nitric oxide synthase: Ca2+ mobilization in human pulmonary arterial endothelial evidence for the existence of stabilization/activation factor. Bio cells. Br. J. Pharmacol. 125,1190-1197 chem. Biophys Res. Commun. 234,476-480 64. Perseclini, A., McMillan, K., and Masters, B.S.S. (1995) Inhibi 51. Iwasaki. T., Wakagi, T., Isogai, Y., Tanaka, K., Iizuka, T., and tion of nitric synthase activity by Zn2+ion. Biochemistry 34, Oshima, T. (1994) Functional and evolutionary implications of a 15091-15095 [3Fe-4S]cluster of the dicluster-type ferredoxin from the ther 65. Miller, R.T., Martasek, P., Raman, C.S., and Masters, B.S.S. moacidophilic archaeon, Sulfolobus sp. strain 7. J. Biol. Chem. (1999) Zinc content of Escherichia coli-expressed constitutive 269,29444-29450 isoforms of nitric-oxide synthase. J. Biol. Chem. 274, 14537- 52. Farm, Y.C., Ahmed, I., Blackburn, N.J., Boswell. J.S., Verk 14540 hovskaya, M.L., Hoffman, B.M., and Wikstrom, M. (1995) 66. Rodriguez-Crespo, I., Nishida, C.R., Knudsen, G.M., and Ortiz Structure of CuB in the binuclear heme-copper center of the de Montellano, P.R. (1999) Mutation of the five conserved his cytochrome aa3-type quinol oxidase from Bocillus subtilis: an tidines in the endothelial nitric-oxide synthase ENDOR and EXAFS study Biochemistry, 334,10245-10255 domain: no evidence for a non-heme metal requirement for 53. Wang. J., Stuehr, D.J., and Rousseau, D.L. (1995) Tetrahydro catalysis. J. Biol. Chem. 274, 21617-21624 biopterin-deficient nitric oxide synthase has a modified heme 67. Perry, J.M., Zhao, Y., and Marietta, M.A. (2000) Cu2+ and Zn2+ environment and forms a cytochrome P-420 analog. Biochemis inhibit nitric-oxide synthase through an interaction with the try, 34.7080-7087 reductase domain. J. Biol. Chem. 275, 14070-14076 54. Iwasaki, T., Hon, H., Hayashi, Y., Nishino, T., Tamura. K., One, 68. Crane, B.R.. Arvai, A.S., Ghosh, S., Getzoff, E.D., Stuehr, D.J., S., lizuka. T., Ogura, T., and Esumi, H. (1999) Characterization and Tainer, J.A. (2000) Structure of the Nw-hvdroxy-L-arginine of mouse nNOS2, a natural variant of neuronal nitric-oxide complex of inducible nitric oxide synthase oxygenase dimer synthase produced in the central nervous system by selective with active and inactive pterins. Biochemistry 39, 4608-4621 alternative splicing. J. Biol. Chem. 274, 17559-17566 69. Tierney, D.L., Martasek, P., Doan, P.E., Masters, B.S.S., and 55. Huang. L., Abu-Soud, H.M., Hill, R., and Stuehr, D.J. (1999) Hoffman, B.M. (1998) Location of guanidino nitrogen of L-argin Nitric oxide-generated P420 nitric oxide synthase: character ine substrate bound to neuronal nitric oxide synthase (nNOS): ization and roles for tetrahydrobiopterin and substrate in pro determination by Q-band pulsed ENDOR spectroscopy J. Am. tecting against or reversing the P420 conversion. Biochemistry, Chem. Soc. 120, 2983-2984 38,1912-1920 70. Tierney, D.L., Huang, H., Martasek, P., Masters , B.S.S., Silver 56. Scott, R.A. (1985) Measurement of metal-ligand distances by man, R.B., and Hoffman, B.M. (1999) ENDOR spectroscopic EXAFS Methods Enzymol. 117, 414-459 evidence for the position and structure of NG-hydroxy-L-argin 57. Rich, A.M., Armstrong, R.S., Ellis, P.J., Freeman, H.C., and Lay, ine bound to holo-neuronal nitric oxide synthase. Biochemistry P.A. (1998) Determination of iron-ligand bond lengths in horse 38,3704-3710

J. Biochem.